Randall Stocker
Karen V. S. Hupp
Chapter 6:
Fire and Nonnative Invasive Plants
in the Southeast Bioregion
Introduction_____________________
This chapter identifies major concerns about fire and
nonnative invasive plants in the Southeast bioregion.
The geographic area covered by this chapter includes
the entire States of Louisiana, Mississippi, and Florida;
all except the northernmost portions of Delaware and
Maryland; the foothill and coastal ecosystems of Virginia, North Carolina, South Carolina, Georgia, and
Alabama; and the lower elevation plant communities
of Arkansas, southeastern Missouri, southeastern
Oklahoma, southwestern Tennessee, and eastern
Texas. This area coincides with common designations of the Atlantic Coastal Plain and the Piedmont
(the plateau region between the Atlantic and Gulf of
Mexico Coastal Plain and the Appalachian Mountains).
Soils are generally moist year-round, with permanent
ponds, lakes, rivers, streams, bogs, and other wetlands.
Elevations vary from 2,407 feet (734 m) on Cheaha
Mountain, Alabama, to –8 feet (–2.4 m) in New Orleans,
Louisiana (USGS 2001). Westerly winds bring winter
precipitation to the bioregion, and tropical air from
the Gulf of Mexico, Atlantic Ocean, and Caribbean
Sea brings summer moisture. Southward through this
region the contribution of winter rainfall decreases, as
does the frequency of freezing temperatures. Tropical
conditions occur at the southern tip of Florida. The
percentage of evergreen species and palms (Serenoa
spp., Sabal spp.) increases along this climate gradient
(Daubenmire 1978).
Plant communities within this portion of the temperate mesophytic forest are complex and subject to
a long history of natural and anthropogenic disturbance. Various methods have been used to estimate
the dominant presettlement forest types. Plummer
(1975) reported that pine (Pinus spp.) and post oak
(Quercus stellata) were the dominant trees on historical survey corner tree lists in the Georgia Piedmont,
and Nelson (1957) used soil type to estimate that 40
percent of the Piedmont was dominated by hardwood
species, 45 percent was in mixed hardwood and pine
stands, and 15 percent was predominantly pine. On
the southeastern Coastal Plain, pine savannas may
have covered between two-thirds and three-fourths of
the area (Platt 1999).
Currently, forests include a mosaic of mostly deciduous angiosperms that form a dense canopy of tall trees
with a “diffuse” layer of shorter, shade-tolerant trees,
interspersed with disturbance- (mostly fire-) derived
pine stands (Daubenmire 1978). Vines are common and
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 91
frequently include native grape (Vitis spp.) species.
Large streams often have extensive floodplains and
oxbows, and areas where the water table occurs at or
near the surface year-round usually support stands of
bald cypress (Taxodium distichum). Coastal dunes are
often populated by American beachgrass (Ammophila
breviligulata) from North Carolina northward and by
sea-oats (Uniola paniculata) throughout the region.
Salt water-influenced wetlands occur landward of
coastal dunes and are dominated by a variety of species including inland saltgrass (Distichlis spicata),
needlegrass rush (Juncus roemerianus), smooth
cordgrass (Spartina alterniflora), and saltmeadow
cordgrass (Spartina patens) (Daubenmire 1978).
While nonnative plants can be found throughout
this region, the highest proportion of nonnative plants
is found in southern Florida (Ewel 1986; Long 1974).
Prior to this century’s increase in transport and trade,
the southern Florida peninsula had geographical and
geological barriers to plant species introductions from
the north, and surrounding waters provided barriers to tropical species introductions. More recently,
southern Florida has become especially vulnerable to
nonnative plant invasions because of a large number
of temperate and tropical species introductions for
horticulture (Gordon and Thomas 1997), the proximity
of the introduction pathways to potentially invasible
habitats, and the relatively depauperate native flora
(Schmitz and others 1997). Additional human-caused
changes in hydrology, fire regime, and salinity have
combined to increase the vulnerability of the vast
low elevation freshwater wetlands south of Lake
­Okeechobee (Hofstetter 1991; Myers 1983).
Discussion of fire and nonnative plant interactions
is complicated by the limited number of experimental
field studies, the lack of a complete understanding of
presettlement fire regimes, and the unknown effects of
increasing atmospheric carbon dioxide and nitrogen and
other aspects of climate change (Archer and others 2001).
Complications notwithstanding, a better comprehension
of the factors and forces at work is critical to developing
fire and other habitat management practices that provide a more effective means of achieving ecological and
societal objectives (D’Antonio 2000).
Fire in the Southeast Bioregion
Naturally occurring fires are, and were, common
in this region (for example, Chapman 1932; Harper
1927; Komarek 1964; Platt 1999; Stanturf and others
2002). The Southeast includes locations with some of
the highest lightning incidence levels on earth. Six of
the eight highest lightning-strike rates in the United
States are found in the southeastern region (TampaOrlando, Florida; Texarkana, Arkansas; Palestine,
Texas; Mobile, Alabama; Northern Gulf of Mexico;
and Gulf Stream-East Carolinas).
92
Little is known about “natural” or prehistoric fire
regimes in the Southeast. During the interval between
the retreat of the ice 18,000 years ago and the initial
influence of Native Americans beginning around 14,000
years ago, variations in soil moisture, lightning strikes,
fuel accumulation, and disturbance history likely resulted in a wide range of fire-return intervals, from
as short as one year to as long as centuries. Similarly,
fire severities probably ranged from minor fires in the
understory to stand-replacement events (Stanturf and
others 2002).
Fire frequency and severity were important factors
in the evolution of southeastern plant communities
(Komarek 1964, 1974; Platt 1999; Pyne 1982a; Pyne and
others 1984; Snyder 1991; Van Lear and Harlow 2001;
Williams 1989), and fire contributes to the high diversity of communities such as pine and shrub (pineland)
communities of south Florida (Snyder 1991) and pine
savannas (Platt 1999). Estimates of presettlement fire
regime characteristics are summarized in reviews by
Wade and others (2000) and Myers (2000) for major
vegetation types in the Southeast bioregion.
Substantial evidence from many disciplines supports the contention that fire was widespread prior
to European arrival (Stanturf and others 2002). Fires
induced by native peoples created and maintained open
woodlands, savannas, and prairies (McCleery 1993;
Williams 1989), and kept forests in early successional
plant communities. Native peoples often burned up
to twice a year and extended the fire season beyond
summer lightning-induced fires (Van Lear and Harlow
2001).
After adopting the practices and utilizing the
clearings made by native people, European settlers
influenced fire patterns and plant communities by
expanding areas of agricultural clearing and repeated burning (Brender and Merrick 1950; Stoddard
1962; Williams 1992), maintaining permanent fields
­(Stanturf and others 2002), introducing large herds of
hogs and cattle (McWhiney 1988; Stanturf and others
2002; Williams 1992), and heavily logging coastal pine
forests, bald cypress, and bottomland hardwood stands
(Stanturf and others 2002; Williams 1989). Frequent
anthropogenic burning, in combination with grazing
cattle and feral pigs, eliminated regeneration of pine
and other woody species in large areas (Brender and
Merrick 1950; Frost 1993).
Subsequent land and fire management practices and
policies oscillated between periods of controlled burning and fire exclusion, and varied from place to place
(Brueckheimer 1979; Johnson and Hale 2000; Paisley
1968; Stoddard 1931). This range of fire practices
was not the result of carefully planned and organized
management strategies but instead was a reaction to
political and social influences at a variety of geographical scales, local to regional. Little scientific information
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
Fire can contribute to the establishment and spread
of nonnative invasive plants under some circumstances
(Mack and D’Antonio 1998). Melaleuca (Melaleuca
quinquenervia), for instance, invades fire-cleared
mineral soils in south Florida (Myers 1975).
Fire exclusion has also been blamed for reducing
native species in favor of nonnatives in fire-adapted
communities. For example, Chinese tallow (Triadica
sebifera) invades fresh marshes (Grace 1999), and
Brazilian pepper (Schinus terebinthifolius) invades
subtropical pine habitats (Myers 2000) in the absence
of fire. Exclusion of fire from longleaf pine (Pinus palustris) communities generally results in woody species
overtopping herbs, thicker duff layers, and changes in
nutrient availability that “all favor extrinsic species at
the expense of endemic residents” (Wade and others
2000, page 66).
Nonnative plant invasions can affect fuel and fire
characteristics in invaded communities (Brooks and
others 2004; Chapter 3) and may subsequently reduce
native plant density and diversity. Altered fuel characteristics associated with some invasive species may
result in fires that kill native plants but not fire-resistant invasive species (Drake 1990; Pimm 1984). For
example, cogongrass (Imperata cylindrica) invasions in
Florida sandhills increase biomass, horizontal continuity, and vertical distribution of fine fuels, compared to
uninvaded pine savanna. Fires in stands invaded by
cogongrass have higher maximum temperatures than
fire in uninvaded stands (Lippincott 2000) and may
therefore cause greater mortality in native species
than fires fueled by native species. Melaleuca invasion
can alter the vertical distribution of fuels such that
communities that typically experienced low-severity
surface fires have a greater incidence of crown fire in
invaded communities (Myers 2000). Conversely, Brazilian pepper and Chinese tallow develop dense stands
that suppress native understory grasses, resulting in
lower fine fuel loads than the fire-maintained plant
communities being replaced (Doren and others 1991;
Grace and others 2001). Lower fuel loads may lead to
reduced fire frequency and lower fire severity, which
may favor the fire sensitive seedling stages of the
invasives (Mack and D’Antonio 1998).
Controlled burning is sometimes used in an effort to manage invasive plants in the Southeast.
However, D’Antonio’s review (2000) suggests that
fire-versus-invasives results are highly variable and
depend on fire intensity, time of burning (Hastings
and ­DiTomaso 1996; Parsons and Stohlgren 1989;
Willson and ­Stubbendieck 1997), weather, and the
status of the remaining seed bank (Lunt 1990; Parsons
and Stohlgren 1989). It is also important to note that,
while dormant season fires have been recommended to
control invasive shrubs in grasslands, they may result
in increases in nonnatives (Richburg and others 2001,
review). It has been recommended that, if fire is used
to reduce populations of nonnative invasive plants,
burning should be timed to reduce flowers and/or seed
production, or at the young seedling/sapling stage
(chapter 4). Spot-burning very small populations of
invasive plants has also been recommended as “cheaper
and easier than implementing a prescribed fire” (Tu
and others 2001).
The limited number of replicated, long-term, field
experiments on fire and invasive plants reflects the
very difficult nature of conducting the needed studies. Even where detailed measurements of the effects
of fire on native and nonnative plant species have
been collected, the studies are typically short-term
and do not necessarily reflect longer-term changes
(Freckleton 2004; Freckleton and Watkinson 2001).
More research is needed over longer periods of time
to better understand the relationships between fire
and invasive species in the Southeast bioregion.
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 93
was available, especially in the early years, to inform
the ongoing debate over fire exclusion and controlled
burning (Frost 1993). Intentional burning practices
rarely attempted to mimic presettlement fire conditions (Doren and others 1993; Drewa and others 2002;
Platt 1999; Platt and Peet 1998; Slocum and others
2003) but were conducted mainly for agriculture and
land clearing.
Contemporary objectives of controlled burns in the
Southeast bioregion include hazard fuels reduction,
wildlife habitat improvement, and range management (Wade and others 2000). Increasing numbers of
acres are being burned for ecosystem restoration and
maintenance (Stanturf and others 2002) and to sustain
populations of rare and endangered plants (Hessl and
Spackman 1995, review; Kaye and others 2001; Lesica
1996). Contemporary fire management practices often
strive to recreate presettlement fire regimes, assuming
that this will promote maximum diversity (Good 1981;
Roberts and Gilliam 1995). Because presettlement fire
regimes are not always well understood, however, it
is difficult to design a fire management program to
meet this objective (Slocum and others 2003).
The negative consequences of past fire management
practices have been interpreted as an “ecological
disaster” (Brenner and Wade 2003). Exclusion of
fire from southeastern pine savannas, for instance,
has been blamed for loss of fire-adapted, species-rich
herbaceous ground cover and subsequent increase in
less fire-tolerant native and nonnative woody species
(DeCoster and others 1999; Heyward 1939; Platt 1999;
Slocum and others 2003; Streng and others 1993;
Walker and Peet 1983).
Fire and Invasive Plants in the Southeast
Bioregion
The remainder of this chapter presents information
on the known relationships of fire and invasive plant
species for five major plant habitats: wet grassland,
pine and pine savanna, oak-hickory (Quercus – Carya)
woodland, tropical hardwood forest, and a brief treatment of cypress (Taxodium distichum) swamp (fig. 6-1).
For each habitat except cypress swamp, a summary
is provided of the role of fire and fire exclusion in
promoting invasions by nonnative plant species, fire
regimes changed by plant invasions, and use of fire to
manage invasive plants, with an emphasis on those
species included in table 6-1. The final section presents
general conclusions and emerging issues relating to
fire and invasive species management in the Southeast bioregion. All parts of the Southeast have been
and continue to be affected by management practices
including fire exclusion, controlled burning, or both
(Brenner and Wade 2003; Freckleton 2004).Therefore,
we have made no attempt to make a distinction between “more managed” (for example, pine plantations)
and “less managed” (for example, conservation areas)
ecosystems in this section.
Wet Grassland Habitat_ ___________
Background
The term “wet grasslands” is used here to include
the Everglades region of southern Florida, grassland
Figure 6-1—Approximate distribution of major plant habitats
in the Southeast bioregion. Upland grasslands and palmetto
prairie are not identified here.
94
savannas with cabbage palm (Sabal palmetto) and cypress in Florida (Küchler’s (1964) palmetto prairie and
cypress savanna, respectively), and the coastal grassdominated wetlands from Virginia to Texas (Küchler’s
(1964) northern and southern cordgrass (Spartina
spp.) prairie). Native species in these wetlands include smooth cordgrass dominating tidally flushed
saltmarshes; smooth and gulf cordgrass (Spartina
spartinae), needlegrass rush, pickleweed (Salicornia
spp.), inland saltgrass, saltmeadow cordgrass, and
saltmeadow rush (Juncus gerardii) in less frequently
flooded more inland marshes. Aquatic species in fresh
marshes include pond-lily (Nuphar spp.), waterlily
(Nymphaea spp.), wild rice (Zizania aquatica), cutgrass (Zizaniopsis miliacea), pickerelweed (Pontedaria
cordata), arrowhead (Sagittaria spp.), cattail (Typha
spp.), maidencane (Panicum hemitomon), spikerush
(Eleocharis spp.), and sedges (Carex spp.) (Wade and
others 2000).
While wet grassland communities border many different habitats, some of the smallest non-graminoid
dominated types include 1-to-several acre (0.5-toseveral hectare) hardwood forest sites found within
marsh, prairie, or savanna in southern Florida. Locally
termed “tree islands,” fire in these locations is usually
driven by processes in the adjacent plant community,
and special features related to the spread of fire into
the tree islands from adjacent wet grasslands are
discussed in the “Tropical Hardwood forest” section
of this chapter.
Wet grassland plant communities tend to be flammable and are adapted to an environment of frequent
wet season (summer) fires (Leenhouts 1982; Schmalzer
and others 1991; Wade 1988; Wade and others 1980).
The following information on fire regimes in these
plant communities comes from Wade and others (2000)
and Myers (2000) (see these reviews for more detail).
Presettlement fire regimes in wet grasslands in much
of the coastal region in the Southeast are classified
as stand-replacement types with 1- to 10-year return
intervals (Myers 2000; Wade and others 2000).
Fire behavior differs among wet grassland types
due to differences in flammability of dominant species, which vary with groundwater levels and salinity. Cordgrass communities in coastal salt marshes
tend to be quite flammable, with green tissues of
saltmeadow cordgrass and gulf cordgrass capable of
burning several times during a growing season. Fire
in these communities will carry over standing water.
Flammability in fresh and brackish marshes is more
variable due to considerable plant diversity. Grass
dominated stands generally experience more intense
and continuous fires than forb and sedge dominated
stands with important exceptions, including cattail
and sawgrass stands (Cladium jamaicense) (Myers
2000; Wade and others 2000).
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 95
Ailanthus altissima
Albizia julibrissin
Alliaria petiolata
Dioscorea alata
Dioscorea bulbifera
Dioscorea oppositifolia
Elaeagnus pungens
Elaeagnus umbellata
Euonymus alatus
Euonymus fortunei
Hedera helix
Imperata cylindrica
Lespedeza bicolor
Lespedeza cuneata
Ligustrum spp.
Lolium arundinaceum
Lonicera japonica
Lonicera spp.
Lygodium japonicum
Lygodium microphyllum
Melaleuca quinquenervia
Melia azedarach
Microstegium vimineum
Paspalum notatum
Pueraria montana var. lobata
Schinus terebinthifolius
Solanum viarum
Triadica sebifera
Scientific name
Wet
grassland
N
N
N
N
N
U
N
U
U
U
N
N
N
N
U
U
N
N
U
H
H
H
P
U
U
H
U
H
Common name
Tree-of-heaven
Mimosa
Garlic mustard
Water yam
Air potato
Chinese yam
Thorny-olive
Autumn-olive
Winged euonymus
Winter creeper
English ivy
Cogongrass
Lespedeza
Sericea lespedeza
Privet species
Tall fescue
Japanese honeysuckle
Bush honeysuckles
Japanese climbing fern
Old World climbing fern
Melaleuca
Chinaberry
Japanese stiltgrass
Bahia grass
kudzu
Brazilian pepper
tropical soda apple
Chinese tallow
N
N
N
N
N
U
N
H
U
U
N
L
H
L
N
H
H
H
U
N
N
H
H
L
U
P
H
H
N
N
N
N
N
N
N
N
N
N
N
U
N
L
U
N
N
N
U
H
H
H
P
U
U
H
U
H
L
U
P
L
H
U
U
U
U
U
U
H
L
L
L
U
H
U
H
H
H
U
U
L
H
H
P
P
H
P
P
P
H
P
P
U
P
P
P
L
L
L
H
N
H
L
P
N
N
P
H
L
H
N
U
P
N
N
N
P
H
U
N
N
N
N
U
N
N
L
N
N
N
U
U
H
H
U
U
N
U
P
N
U
Tropical
hardwood
forest
Forest and woodland
Pine and pine Oak-hickory
Upland grassland Palmetto prairie
savanna
woodland
Grassland
N
N
N
L
H
U
N
N
N
N
U
N
N
N
N
N
N
N
U
H
H
N
U
N
U
P
N
U
Cypress
swamp
Table 6-1—Major plant habitats in the Southeast bioregion and perceived threat potential of several important nonnative plants in each habitat (L= low threat, H = high threat, P = potentially high
threat, N= not invasive, U = unknown). Designations are approximations based on available literature.
Postfire succession patterns in wet grasslands are
influenced by season of burning and the interplay
of hydroperiod and fuels, which together determine
whether fires are lethal or nonlethal to the dominant
species. Hydroperiod factors that influence the effect of
fire on belowground plant parts and substrate include
proximity to the water table, tidal conditions, and
drought cycles. In most cases, the aboveground vegetation is consumed by fire, and the dominant species
that make up the fuel sprout from underground buds,
tubers, or rhizomes after fire. Peat fires and postfire
flooding are two disturbance events that can kill both
above- and belowground organs of existing vegetation.
Severe peat fires can occur in organic substrates when
severe drought coincides with low water table levels.
Vegetation can also be killed by water overtopping
recovering vegetation after a fire. Successional species
would be expected to be primarily those represented in
the seed bank (Myers 2000; Wade and others 2000).
In some areas, native wet grassland communities
have been altered by fire exclusion, allowing invasion
of woody species. Fire is now being reintroduced in
many areas to restore native species compositions (for
example, see Leenhouts 1982).
Role of Fire and Fire Exclusion in
Promoting Nonnative Plant Invasions in
Wet Grasslands
Fire exclusion in wet grasslands during the past
century has resulted in less frequent but more severe
fires than occurred prior to European settlement. These
fires have opened up wet grasslands to invasion by
nonnative plant species (Bruce and others 1995).
Melaleuca is well adapted to survive fire and to establish and spread in the postfire environment. There
may not be many better fire-adapted tree species in
the world than melaleuca. Nicknamed the “Australian
fireproof tree” (Meskimen 1962), its native habitats
include fire-shaped ecosystems in Australia (Stocker
and Mott 1981). It produces serotinous capsules (le
Maitre and Midgley 1992) that release as many as 20
million seeds per tree (Woodall 1981) following fire.
Complete capsule dehiscence can occur as quickly
as a few days following a crown fire (Woodall 1983).
Seedlings establish on fire-cleared mineral soil and
can survive fire within a few weeks or months after
germination (Meskimen 1962; Myers 1975, 1983).
Melaleuca sprouts from roots and from epicormic
trunk buds to resume growth after fire. Although the
outer bark can easily burn (fig. 6-2), the trunk wood is
protected from fire damage by spongy inner bark that
is saturated with water (Turner and others 1998).
Fire is not necessary for melaleuca establishment
(Woodall 1981); in fact, melaleuca spreads readily
with changes in hydrology and mechanical damage
96
to habitats (Cost and Carver 1981). Nevertheless, its
ability to capitalize on burned areas is remarkable
(Hofstetter 1991; Myers 1983). “If melaleuca were
managed as a desired species, prescribed fire would be
the single most important tool available to the resource
manager” (Wade 1981).
Melaleuca seedlings that establish after fire at the
height of the dry season may have 5 to 8 months’ advantage over native tree species such as south Florida
slash pine (Pinus elliottii var. densa) and bald cypress,
which release seed during the wet season (Wade 1981).
Other species that may establish and/or spread following fire in wet grasslands include climbing ferns
(Lygodium spp.) and the shrub chinaberry (Melia
azedarach). According to a review by Ferriter (2001),
Old World climbing fern (Lygodium microphyllum)
occurs in sawgrass marsh in southern Florida and
may spread following fire. No published information
was found related to the response of chinaberry to fire
in the Southeast; however, it reproduces vegetatively
from both stumps and roots following fire in Argentina (Menvielle and Scopel 1999; Tourn and others
1999).
In areas where fire has been excluded from wet
grassland communities, native species are often
Figure 6-2—Melaleuca’s papery trunk. (Photo by Forest & Kim
Starr, United States Geological Survey, Bugwood.org.)
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
replaced by a dense woody overstory composed of
nonnative invasive trees and/or shrubs (for example, melaleuca, Chinese tallow, Brazilian pepper
and chinaberry). When this occurs, native species
numbers and diversity are dramatically reduced
(for example, Bruce and others 1995). It has been
suggested that fire exclusion can contribute to the
invasion of Chinese tallow into coastal prairies
(Bruce and others 1995; chapter 7; D’Antonio 2000).
Similarly, reduced fire frequency due to humaninduced changes to hydrology is blamed for invasion
of Brazilian pepper into sand cordgrass (Spartina
bakeri) and black rush (Juncus roemerianus) dominated salt marshes in Florida’s Indian River Lagoon
(Schmalzer 1995).
The relationship between fire and fire exclusion
and the increase in Brazilian pepper in south Florida
wetlands is not well understood. Brazilian pepper
is not a fire-adapted species (Smith, C. 1985) and is
generally kept out by fire in adjacent pinelands (Loope
and Dunevitz 1981). The response to fire in wetlands
may not be the same as in pinelands, although the
field studies have not been conducted on “typical”
wetland plant communities. The limited published
research includes assessment of the effect of repeated
fire (generally every two years) on experimental plots
in highly altered “rock-plowed” limestone substrate.
Much of this formerly agricultural area was originally
sawgrass marsh, although many woody species invaded the rock-plowed portions after the fields were
abandoned. Fire exclusion (control plots in this study)
resulted in increased Brazilian pepper stem density,
but Brazilian pepper stem density also increased in
burned plots (Doren and others 1991). The effects of
fire were related to the size of the Brazilian pepper
plant. Smaller, apparently younger plants were badly
damaged or killed by fire, while larger plants either
recovered completely or did not burn. The conclusion
of the authors was that Brazilian pepper invasion
progressed with or without fire, and that fire is not
an appropriate management tool for this unique area
(Doren and others 1991). Dry season wildfires are
thought to contribute to increasing Brazilian pepper
populations in “tree islands,” which are typically treedominated areas within the wet grasslands of southern
Florida (Ferriter 1997, FLEPPC review) (see “Tropical
Hardwood Forest Habitat” section page 107).
Chinaberry may also invade wet grasslands where
fire has been excluded. This shrub occurs primarily
in disturbed areas but is also said to invade relatively
undisturbed floodplain hammocks, marshes, and upland woods in Florida (Batcher 2000b, TNC review),
although no additional information is available. In
Texas, riparian woodlands and upland grasslands have
also been extensively invaded by chinaberry (Randall
and Rice unpublished, cited in Batcher 2000b).
Tall fescue (Lolium arundinaceum) is found on disturbed upland grassland sites or where “the natural
fire regime has been suppressed (Eidson 1997)” (as
cited in Batcher 2004, TNC review).
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 97
Effects of Plant Invasions on Fuel and Fire
Regime Characteristics in Wet Grasslands
Several nonnative invasive plants are thought to
change fire regimes in wet grasslands in the Southeast
bioregion by changing the quantity and/or quality of
fuels in invaded communities. Changes in fuels may
subsequently reduce or increase fire frequency and
severity. Examples of both cases are evident in wet
grassland plant communities.
Observational and limited experimental evidence
suggests that invasive hardwoods such as Chinese
privet (Ligustrum sinense), Chinese tallow, and Brazilian pepper shade out and/or replace native plant species
in southeastern marshes and prairies such that fine
fuel loads and horizontal continuity are reduced (for
example, Doren and Whiteaker 1990; Doren and others
1991; Grace 1999; Platt and Stanton 2003). When this
occurs, fire frequency and intensity may be reduced
and fire patchiness increased. However, there is little
experimental evidence to support these conjectures,
and more research is needed to better understand
the implications of these vegetation changes on fire
regimes.
Melaleuca invasion can have variable effects on
fuels and fire behavior. Large amounts of litter under
melaleuca stands (Gordon 1998) promote intense and
severe fires (Flowers 1991; Timmer and Teague 1991)
that are difficult to control and have large potential
for economic damage, loss of human life and property,
and negative ecological consequences (Flowers 1991;
Schmitz and Hofstetter 1999, FLEPPC review; Wade
1981). These high intensity fires promote melaleuca
establishment and spread, and reduce cover of native
species. Severe fire also removes the outer, highlyflammable melaleuca bark layers, thus reducing the
probability of damage to mature melaleuca from subsequent fires (Wade 1981). Thus, a positive feedback loop of
fire-promoting-melaleuca and melaleuca-promoting-fire
is created (Hofstetter 1991; Morton 1962). Conversely,
intense fires fueled by melaleuca may reduce the chances
of subsequent fires when organic soils are consumed
and the elevation of the soil surface lowered (Schmitz
and Hofstetter 1999). Small changes in water level can
then theoretically reduce the likelihood of fire by flooding
these formerly unflooded sites.
Old World climbing fern alters plant community
fuel structure in wet grasslands and associated tree
islands with extensive, dry-standing frond “skirts”
that create ladder fuels that facilitate fire spread
into tree canopies (Ferriter 2001, FLEPPC review)
(see section on “Tropical Hardwood Forest Habitat”
page 107). Roberts (D. 1996) also reports that fire
penetrates into wet grasslands from the margins of
forested communities where Old World climbing fern
has invaded and provides a novel source of additional
fuel (see section on “Pine and Pine Savanna Habitat”
page 100). Fire spread may also be promoted by pieces
of burning fern frond blowing aloft into grasslands
from tree islands (Roberts, D. 1996).
While the potential exists for invasive plant species
to influence abiotic factors that affect fire behavior,
including water table elevation and surface hydrology,
this relationship has not yet been shown to be important in wet grasslands in the Southeast bioregion. It
is logical to assume, for example, that invasive species
that lower the water table through evapotranspiration could reduce soil moisture and thus affect subsequent fire characteristics. It has been suggested that
melaleuca increases the amount of water lost to the
environment in sites with standing water by adding its
evapotranspiration to the water surface evaporation
(for example, Gordon 1998; Schmitz and Hofstetter
1999; Versfeld and van Wilgen 1986; Vitousek 1986).
In the southwestern United States, transpiration of
dense stands of nonnative tamarisk (Tamarix spp.) can
result in the loss of large quantities of water on sites
where the water table is just below the soil surface
(Sala and others 1996). In many wet grasslands in
the southeastern United States, however, the water
table is at or above the soil surface, and evaporation
and evapotranspiration are both driven and limited
by solar energy. Simply adding another species to the
system does not increase the available energy and
therefore does not increase the amount of water lost
to the atmosphere, although it may increase the available evaporative surface (Allen and others 1997).
Other invasive species-induced changes in hydrology
may have some effect on wet grasslands in the Southeast. The thick (over 1 m) rachis mat formed by decades
of Old World climbing fern growth may have diverted
shallow stream meandering of the Loxahatchee River
in east-central Florida by a distance of about 164 feet
(50 m) (R. Stocker, personal observation, fall 1997).
At this scale only very small portions of the invaded
habitat would be affected. Additional study of the relationships among invasive plants and abiotic factors
that affect fire regime is warranted.
Use of Fire to Manage Invasive Plants in
Wet Grasslands
Controlled burning has been used extensively to
manage invasive plants in the Southeast bioregion,
with varied results. Only a small portion of the literature describes research on the use of fire to control
invasives in wet grasslands.
98
Many wet grassland sites are on organic soils, and
fires occurring when the organic surface soil is dry can
consume the peat and affect the type of vegetation that
subsequently develops on the site (Ferriter 2001; Myers
2000; Schmitz and Hofstetter 1999). Therefore, it may
be possible to prescribe fires that could substantially
damage plant roots in wet grasslands during dry periods (Nyman and Chabreck 1995). Documented success
using such burns to control invasive nonnative plant
species, however, is lacking (Wade and others 2000),
and care must be taken to avoid substantial damage
to desirable species.
Frequent fires in wet grasslands during historic and
prehistoric times are thought to have maintained grasslands with very little woody vegetation (Schmalzer
1995). It follows that prescribed fires with a frequency
and seasonality within the reference range of variation experienced in these habitats might favor native
wet grassland species over nonnative woody species.
Controlled burning following flooding or plant flowering has been suggested as particularly effective in
reducing “unwanted woody vegetation” in salt marshes
of the St. Johns National Wildlife Refuge (Leenhouts
1982). Fire is not, however, effective for controlling all
invasive species in wet grasslands, some of which are
well adapted to frequent fires.
Controlled burning has been promoted as a means to
reduce woody vegetation in salt marshes (Leenhouts
1982) but has not been effective in controlling melaleuca (Belles and others 1999, FLEPPC review) and
has provided mixed results for Chinese tallow (Grace
1999; Grace and others 2001) and chinaberry (Tourn
and others 1999).
Controlled burning alone is not effective at controlling
melaleuca (Wade 1981) and will not eliminate mature
stands (Belles and others 1999). Fires timed to consume
seedlings after most germination has occurred have the
best potential to control melaleuca (Woodall 1981). Fire
can kill melaleuca seedlings less than 6 months old (Belles
and others 1999); however, it is difficult to achieve the
needed degree of soil surface dryness and fuel load to
carry a fire severe enough to prevent postfire sprouting.
Melaleuca seedlings less than 1 year old may sprout from
root collars after fire damage (Myers 1984). Susceptibility to fire-induced mortality is reduced as seedlings and
saplings grow taller, with more than 50 percent of 1.5foot-(0.5 m) tall saplings surviving in one study (Myers
and others 2001). Because melaleuca seeds are able to
survive in flooded organic soils for about 1.5 years and in
unflooded sandy soils for 2 to 2.3 years (Van and others
2005), postfire establishment from the soil seed bank is
also a concern.
Some resource managers maintain that melaleuca
control can be achieved with proper timing of prescribed
burns (Belles and others 1999; Maffei 1991; Molnar
and others 1991; Pernas and Snyder 1999), but the
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
success of this approach depends on postfire rainfall,
which often does not follow anticipated patterns. Two
seasonal windows of opportunity may exist, depending
on rainfall patterns (Belles and others 1999). (1) Burning during the late wet season, when surface soils are
likely to be moist but not flooded, would encourage
melaleuca seed germination just prior to soil dry-down
during the dry season. With average dry-season rainfall, melaleuca seedlings are likely to die before the wet
season returns the following May or June. If dry-season
rainfall is above average, however, melaleuca recruitment is likely to be high. (2) Burning at the beginning
of the wet season also encourages seed germination,
and normal rainfall patterns might provide sufficient
flooding to kill seedlings. Fluctuating rainfall patterns
or less than average quantity during the wet season
could result in substantial melaleuca recruitment
(Belles and others 1999). Because melaleuca has very
small wind- and water-dispersed seeds, reproductive and
outlying individuals must be killed if long-term reduction
of populations is to be achieved (Woodall 1981).
Repeated fires may have potential for controlling
melaleuca; however, fuel loads may be insufficient to
carry fire in consecutive years. Some wet grasslands
might be capable of providing sufficient fuel for a second fire within 2 or 3 years after the first fire. Nearly
all melaleuca seedlings were killed in a second fire
2 years after a wildfire in a wet grassland dominated
by muhly grass (Muhlenbergia capillaris) (Belles and
others 1999).
Recommendations for controlling mature melaleuca
stands include using fire only after first killing reproductive individuals with herbicide (Myers and others
2001) (fig. 6-3). Herbicide-treated trees release large
quantities of viable seed. Prescribed burning should
then be conducted within 2 years (6 to 12 months
recommended for Big Cypress Preserve; Myers and
others 2001) of the herbicide-induced seed release
and subsequent germination (Belles and others 1999).
Mature melaleuca stands burned by wildfire should
have high priority for management because of the
potential for postfire spread following seed release
from fire-damaged melaleuca or adjacent stands of
unburned melaleuca. Recommendations include additional specifications for herbicide use and careful
monitoring for several years after fire (Belles and
others 1999).
Repeated burning, especially in combination with
other control methods, can effectively control Chinese
tallow under some circumstances. Chinese tallow is
difficult to manage with fire because fuel loads under
tallow infestations are often insufficient to carry fire
(Grace 1999). See chapter 7 for more information on
the use of fire to control Chinese tallow.
The effect of repeated fire (generally every two
years) to control Brazilian pepper has been evaluated
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 Figure 6-3—Burning herbicide-killed melaleuca at South
Florida Water Management District. Melaleuca stand was
treated with herbicide in spring of 1996 and burned in winter
2001. The objective was to consume a large portion of the
standing dead biomass with the fire, but only the tops of
trees were burned. (Photo by Steve Smith.)
in highly altered “rock-plowed” limestone substrate
in south Florida. Fine fuel supply was insufficient to
carry annual fires, which the authors attributed to
the replacement of graminoid species with Brazilian
pepper (Doren and others 1991). While density and
coverage of Brazilian pepper had increased on both
burned and unburned plots at the end of the 6 years
of evaluation, increases on burned plots occurred more
slowly than on unburned plots (Doren and others 1991).
Control of Brazilian pepper using herbicide followed by
prescribed burning has also been attempted (fig. 6-4),
though results are not reported in the literature.
Figure 6-4—Using a helitorch to ignite herbicide-killed
Brazilian pepper in spring 2006. Herbicide was applied
one year before the fire. The fire was not effective at
consuming dead pepper trees due to standing water and
low fuel loads. (Photo by Steve Smith.)
99
Published reports were not found that document attempts to manage chinaberry with fire in the Southeast
bioregion. Chinaberry recovered fully from a single autumn fire in South America, reproducing vegetatively
from both stumps and roots (Menvielle and Scopel 1999;
Tourn and others 1999). A single surface fire killed all
seeds in the seed bank, and fruit production was 90
percent less than in unburned control plots. Chinaberry
seedling emergence following the fire was 5 to 20 times
greater in unburned control plots than in burned plots;
however, the seasonal pattern of seedling emergence
and survivorship was not affected by fire (Menvielle and
Scopel 1999). The South American studies suggest that
a single fire is not effective in controlling chinaberry,
with populations quickly returning to prefire levels or
even expanding (Tourn and others 1999). Additional
research is needed to determine if fire in different
seasons, multiple-year fires, or a combination of fire
and herbicide application are effective for controlling
Chinaberry in the Southeast.
The use of fire alone is not likely to cause enough
damage to kill Old World climbing fern plants and
prevent postfire sprouting and rapid recovery in wet
grasslands. This is due to the high moisture content
of most wet grassland fuels resulting in low-severity
fire (Stocker and others 1997). Spot burning prior
to herbicide application can reduce the amount of
herbicide needed to control Old World climbing fern
by about 50 percent (Stocker and others, In press).
See the “Pine and Pine Savanna Habitat” section for
more information on the use of fire to control climbing
ferns.
Autumn-olive (Elaeagnus umbellata), sericea
­lespedeza (Lespedeza cuneata), shrubby lespedeza
(L. bicolor), and tall fescue commonly occur on disturbed
sites near southeastern grasslands. While no specific
studies are available that examine the relationship
between these species and fire in Southeast grassland
habitats, studies conducted in other areas suggest that
prescribed fire has a limited potential for controlling
these species under certain circumstances and in
combination with other control methods. Autumnolive may respond to fire damage by sprouting, but
empirical information on the relationship of this species to fire and fire management is lacking (Munger
2003b, FEIS review). See FEIS reviews by Munger
(2004) and Tesky (1992) and TNC reviews by Stevens
(2002) and Morisawa (1999a) for more information on
the use of fire for management of lespedeza species.
Recent introductions of tall fescue can be controlled
by spring burning, and combinations of prescribed
burns and herbicide applications have “moderate to
high potential for restoration” (Batcher 2004).
Several resource management organizations have
suggested that fire can be used successfully to reduce
south Florida silkreed (Neyraudia reynaudiana)
100
­ opulations prior to spraying regrowth with herbip
cide. They caution, however, that silkreed is a highly
combustible fuel source and, because of that, a special
burning permit may be required (Rasha 2005, review).
Burning without follow-up herbicide or mechanical
control is ineffective in controlling silkreed and may
enhance its growth and spread (Guala 1990, TNC
review).
Other species that can be found in drier or upland
portions of the wet grassland habitat include Japanese
honeysuckle (Lonicera japonica), bush honeysuckles
(Amur honeysuckle (L. maackii), Morrow’s honeysuckle
(L. morrowii), and tatarian honeysuckle (L. tatarica),
Japanese stiltgrass (Microstegium vimineum) and
tropical soda apple (Solanum viarum) (table 6-1).
Bush honeysuckles are typically top-killed by fire,
and fire may kill seeds and seedlings. Adult plants
probably survive by postfire sprouting from roots and/
or root crowns. Studies conducted in the Southeast
bioregion are not available; however, field work in the
Northeast bioregion suggests that repeated prescribed
fire may be useful in controlling bush honeysuckles
(chapter 5; Munger 2005a, FEIS review). Fire research
on Japanese stiltgrass in the Southeast bioregion is
also needed; however, studies outside of the Southeast
suggest that prescribed fire prior to seed set might
aid in controlling this species (Howard 2005c, FEIS
review). No published information is available on the
relationship of tropical soda apple to fire.
Pine and Pine Savanna Habitat_ ____
Background
Pine and pine savanna habitats covered here include
southern mixed forest, oak-hickory-pine forest, and
subtropical pine forest associations as described by
Küchler (1964). Pine and oak species are the dominant
trees, including longleaf pine, shortleaf pine (Pinus
echinata), loblolly pine (P. taeda), slash pine (P. elliottii), pond pine (P. serotina), southern red oak (Quercus
falcata), turkey oak (Q. laevis), sand-post oak (Q.
margaretta), bluejack oak (Q. incana), blackjack oak
(Q. marilandica), post oak, and water oak (Q. nigra).
Pond cypress and palms are the dominant trees in
some wetter and more southern sites. Shrub species
are common, including runner oaks (Q. minima and
Q. pumila), sumac (Rhus spp.), ericaceous shrubs (for
example, Vaccinium), palms, wax myrtle (Myrica cerifera), and hollies (Ilex spp.). Understory species include
grasses such as wiregrass (Aristida stricta and A. beyrichiana), little bluestem (Schizachyrium scoparium),
and numerous forbs. When Europeans first arrived
in the Southeast, pine stands, and especially pine
savannas, may well have been the dominant vegetation in most of this area, extending from southeastern
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
Virginia to eastern Texas and from northern Georgia
and Alabama to the Florida Keys (Platt 1999).
Presettlement fire regimes are poorly understood,
but it is inferred that the high number of lightning
strikes resulted in a fire-return interval of less than 13
years in pine forests and savannas. Larger and more
intense fires probably occurred in May and June, after
the start of the lightning/rain season but before large
amounts of rain had fallen. Summer fires were probably
more frequent but less intense and smaller in area.
Ignitions by Native Americans probably increased fire
frequency in many locations, shaping the savannas
seen by early explorers (Wade and others 2000). Fire
intervals may have been 1 to 4 years (1 to 5 years for
subtropical pine forest; Myers 2000) before the arrival
of European settlers, and then 1 to 3 years until fire
exclusion became the norm in the early 1900s (Wade
and others 2000).
Fires in these habitats were historically understory
fires. Short return-interval (<10 years), understory fires
predominated in most of the southern mixed forest and
oak-hickory-pine types (sensu Küchler 1964). Slash
pine and loblolly pine habitats experienced understory
and mixed-severity fire regimes, with presettlement
fire-return intervals estimated between 1 and 35 years
(Myers 2000; Wade and others 2000).
The once-common southern pine forests were dramatically reduced by invasions of native hardwood
species when fire exclusion policies were adopted in the
1920s and 1930s. Also affected were the populations of
native plant and wildlife species, nutrient cycling, fuel
reduction, and range management objectives that were
associated with the historical fire regime. It has been
suggested that savanna ecosystems that are not too
seriously degraded can be restored if the appropriate
fire regime (short return-interval, understory, spring
and summer fires) is re-introduced, because the native
plant species are adapted to this regime (Wade and
others 2000).
Among the most threatening nonnative invasive
plant species found in these habitats are cogongrass,
Japanese honeysuckle, Brazilian pepper, and melaleuca. These species are fire-tolerant, thus reducing the
effectiveness of fire in controlling their establishment
and dispersal. Additionally, populations of climbing
ferns appear to be increasing rapidly in Florida pine
habitats and are spreading in Alabama, Florida, Georgia, and Mississippi pine habitats. Japanese climbing
fern (Lygodium japonicum) has become particularly
troublesome where pine straw is collected for sale as
mulch. Spores of this fern have been found in straw
bales, and the distribution of mulch bales throughout
the Southeast has spread Japanese climbing fern into
new areas. There is no specific information on fire and
management of this species in this vegetation type. Old
World climbing fern and melaleuca are also ­serious
problems in wet grasslands and are discussed in the
“Wet Grasslands Habitat” section.
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 101
Role of Fire and Fire Exclusion in
Promoting Nonnative Plant Invasions in
Pine and Pine Savanna
Frequent surface fires typical of presettlement fire
regimes in most pine habitats promote some invasive
species such as cogongrass. When fire is excluded,
pine habitats are especially vulnerable to invasions
by nonnative plants such as Brazilian pepper and
Japanese honeysuckle.
Cogongrass and closely related Brazilian satintail
(Imperata brasiliensis) are perennial, rhizomatous
grasses that are well adapted to frequent fire. Both
are early-seral species in wet-tropical and subtropical
regions around the world. Discussion about management of these species is complicated by difficulty in
distinguishing between the two species, lack of consensus among taxonomists about whether they actually are separate species, and determination of their
native ranges (Howard 2005b, FEIS review). Among
the more recent treatments, Wunderlin and Hansen
(2003) describe them as distinct species, distinguished
by anther number, and suggest that, while cogongrass
is not native to the United States, Brazilian satintail
is native to Florida.
Cogongrass requires some type of disturbance, such
as fire, to maintain its dominance in southeastern pine
understory. Frequent fire typically favors cogongrass
over native species including big bluestem (Andropogon
gerardii), Beyrick threeawn (Aristida beyrichiana),
golden colicroot (Aletris aurea), and roundleaf thoroughroot (Eupatorium rotundifolium). Cogongrass
flowering and seed production may be triggered by
burning and other disturbances, although flowering
has also been observed in undisturbed populations. In
the absence of fire, vegetative growth from rhizomes
(fig. 6-5) allows expansion of populations. Rhizomes
also sprout easily after burning (Howard 2005b). In
Mississippi wet pine savannas, cogongrass seedlings
had higher levels of survival (for at least two months)
in burned than in unburned study plots (King and
Grace 2000).
Only limited information is available on two invasive
shrub species. Brazilian pepper invades pine rockland
(southern Florida habitat on limestone substrate)
where fire has been excluded (Loope and Dunevitz
1981). It is suggested that Japanese honeysuckle is
intolerant of frequent, low-severity fire and is typically
absent from plant communities with this type of fire
regime, such as longleaf pine. Therefore exclusion of
fire from these communities may promote its establishment and spread (Munger 2002a, FEIS review).
Figure 6-5—Cogongrass rhizomes. (Photo by Chris Evans,
River to River CWMA. Bugwood.org.jpg.)
Old World climbing fern invasions provide a novel
source of fuel in pine habitats (Roberts, D. 1996) and
alter fire behavior by altering plant community fuel
structure with extensive dry-standing frond “skirts”
that ladder fire into the canopies of trees (Ferriter 2001)
(fig. 6-7). Resulting canopy fires often kill trees that
are adapted to low-severity surface fires, as well as
native bromeliads (for example, wild pine (Tillandsia
fasciculata)) resident on tree trunks. Fire spread may
also be promoted when pieces of burning fern frond are
kited into adjacent areas (Roberts, D. 1996). Increased
fuel loads, altered fuel structure, and spotting from
Old World climbing fern are blamed for tree mortality and escape of prescribed fires in pine stands at
Jonathan Dickinson State Park in Florida. The park’s
fire management plan has been revised to no longer
depend on wetland buffers to act as fire breaks if they
contain climbing fern (Ferriter 2001).
Effects of Plant Invasions on Fuel and Fire
Regime Characteristics in Pine and Pine
Savanna
Nonnative species life-forms that have invaded
southeastern pine habitats and altered fuel and fire
regime characteristics include trees (melaleuca),
shrubs (Brazilian pepper), grasses (cogongrass), and
ferns (climbing ferns) (fig. 6-6). In some situations
these species replace native species and fill similar
forest strata, but in other cases the invasive plants
completely alter the horizontal structure and fuel
characteristics of the invaded plant community.
Figure 6-6—Old World climbing fern climbing and overtopping vegetation in a pine habitat at the Jonathan Dickinson
State Park in central Florida. (Photo by Mandy Tu, The Nature
Conservancy.)
102
Figure 6-7—Old World climbing fern on a slash pine, burning
during a routine prescribed burn in pine flatwoods at the Reese
Groves Property in Jupiter, Florida. Large slash pine and cypress
have been killed by fire when climbing fern “ladders” carry fire
into the canopy. (Photo by Amy Ferriter, South Florida Water
Management District, Bugwood.org.)
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
Cogongrass invasion changes fuel properties in southeastern pine communities (Howard 2005b) (fig. 6-8).
In a study of fuels and fire behavior in invaded and
uninvaded pine stands in Florida, Lippincott (2000)
found that invasion by cogongrass may lead to changes
in fire behavior and fire effects in these communities.
Native plant and cogongrass fuels have similar energy
A
B
Figure 6-8—(A) Infestation of cogongrass in a slash pine plantation in Charles M. Deaton Preserve, Mississippi. (Photo by
John M. Randall, The Nature Conservancy.) (B) Cogongrass
among planted pines in Mitchell County, Georgia, forms large
accumulation of fine fuels around the base of trees. (Photo by
Chris Evans, River to River CWMA, Bugwood.org.)
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 content; however, fire behavior is driven by factors
other than energy content of fuels. Sites invaded by
cogongrass had greater fine-fuel loads, more horizontal
continuity, and greater vertical distribution of fuels.
The resulting fires were more horizontally continuous
and had higher maximum temperatures and greater
flame lengths than fires in adjacent plots not invaded
by cogongrass, and they resulted in higher subsequent
mortality to young longleaf pine (Lippencott 2000).
Similarly, Platt and Gottschalk (2001) found fine
fuel and litter biomass were higher in cogongrass
and nonnative silkreed stands than in adjacent pine
stands without these grasses. The authors suggest that
increases in fine fuels attributed to cogongrass could
increase fire intensity at heights of 3 to 7 feet (1 to 2
m) above the ground (Platt and Gottschalk 2001).
Bahia grass (Paspalum notatum) is not commonly
found within existing pine stands but occupies heavily disturbed pine habitat that has been converted to
pasture and rangeland, and interferes with efforts to
restore pine communities. Bahia grass forms a continuous “sod fuel layer” in the previously patchy pine
community, thus increasing fuel continuity (Violi 2000,
TNC review). Because bahia grass is important forage
for livestock, rangeland managers use controlled burns
in winter to stimulate its growth. Winter burns negatively affect some native understory species, including
wiregrass, which responds better to late summer and
fall burns (Abrahamson 1984).
Melaleuca invasion in pine flatwoods can alter the
fire regime from frequent (1- to 5-year return interval),
low-severity surface fires to a mixed regime with less
frequent (<35 to 200 year return interval) fires and
greater incidence of crown fires. Crown fires are typically nonlethal to melaleuca trees but usually result
in pine mortality. This combination of high-intensity
fire and crown-fire survival is uncommon in North
America (Myers 2000).
Low levels of fuel under mature Brazilian pepper
(Doren and others 1991) and the difficulty of burning Brazilian pepper wood and leaves due to their
high moisture content (Meyer 2005a, FEIS review)
probably reduce fire intensity and fire spread in
areas of dense infestation. Similarly, invasion by
kudzu (Pueraria montana var. lobata) may reduce
flammability of invaded pine habitats during the
growing season due to its luxuriant, moist foliage.
Conversely, the large amount of fuel biomass contributed by kudzu (fig. 6-9) and by plants killed by
its invasion may increase the potential for dormantseason fires by increasing fuel loads, and its vining
nature may increase the chance of fire crowning
(Munger 2002b, FEIS review). These conjectures
have not, however, been tested empirically.
103
A
B
Figure 6-9—Kudzu infestation at Travelers Rest, South Carolina
in (A) summer, and (B) winter. Green kudzu foliage in summer
may reduce potential for fire spread, while the opposite may be
true in winter. (Photos by Randy Cyr, GREENTREE Technologies, Bugwood.org.)
Use of Fire to Manage Invasive Plants in
Pine and Pine Savanna
Fire has become an important tool of natural area
managers for removal of nonnative invasive species and
maintenance of fire-adapted pine communities (Rhoades
and others 2002). A study in a pine flatwoods community
found that annual winter (non-growing season) burning
increased native species richness and provided habitat for
rare and listed plant species. Nonnative invasive species
were found only in unburned plots. This may, however,
be attributed to microsite differences within the treatment areas, as there was greater moisture availability in
burned plots than in unburned plots (Beever and Beever
1993).
Old World climbing fern aerial fronds burn easily,
and individual fronds can be completely consumed
by a fire of sufficient intensity, but there are several
reasons why prescribed fire is not expected to provide
104
a major role in management of this species (Ferriter
2001). Pieces of burning climbing fern fronds get caught
in fire-induced updrafts, reducing the ability to manage the fire perimeter when they are transported to
adjacent areas. Climbing fern spores are very small
and probably travel great distances by wind, including fire currents. Old World climbing fern plants in a
south Florida slash pine stand were observed to sprout
and recover rapidly after low-severity fire was applied
using a hand-held propane torch (Stocker and others
1997).
Brazilian pepper is another species that is unlikely
to be eliminated from pine stands by fire. Low-severity
fire does not kill adult pepper trees, as girdling of the
stem results in profuse sprouting from aboveground
stems and root crowns (Woodall 1979). Brazilian pepper
seeds can be killed by heat (70 °C for 1 hour; Nilsen
and Muller 1980), and young seedlings can be killed
by fire (Ferriter 1997). However, the intense crown
fires necessary to kill adult plants (Doren and others
1991; Smith, C. 1985) do not commonly occur in pine
stands with dense Brazilian pepper infestation. Fire
does not carry well in mature Brazilian pepper stands,
and fire rarely penetrates dense stands (Meyer 2005a).
Brazilian pepper litter decomposes rapidly, leaving
little litter for fuel, and moisture levels of branches,
leaves, and litter are typically high (Doren and others
1991).
Prescribed fire may be more effective for controlling
young Brazilian pepper stands. In areas where the
water table lies below the soil surface for at least part
of the year, grasses should provide sufficient fuels to
carry fire of sufficient severity to kill young Brazilian
pepper seedlings, and may also kill seeds (Nilsen and
Muller 1980). Maintaining fire programs that killed
seedlings prior to reaching unspecified “fire-resistant
heights” has resulted in pepper-free areas (Ferriter
1997), and it has been noted that fire with a 5-year
fire-return interval in Everglades National Park
has excluded Brazilian pepper (Loope and Dunevitz
1981). On sites where either higher or lower water
tables reduce the development of herbaceous fuels,
prescribed fire may not be of sufficient severity to kill
young Brazilian pepper plants (Ferriter 1997). In a
study in south Florida pinelands, for example, most
Brazilian pepper saplings over 3 feet (1 m) tall survived
fire by coppicing (Loope and Dunevitz 1981). In any
case, Brazilian pepper seed is readily dispersed from
nearby stands by animals (Ewel and others 1982).
The conclusion of a group of resource managers and
scientists is that repeated burning may slow invasions
of this species by killing seeds and seedlings, but fire
“is not an effective control method for mature Brazilian peppertree stands” (Ferriter 1997).
It has been suggested that Japanese honeysuckle can
be controlled by prescribed burning in pine plantations
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
or in fire-dependent natural communities. Prescribed
burns in Virginia are recommended to reduce Japanese
honeysuckle cover and to “inhibit spread” for 1 to 2
growing seasons (Williams 1994, Virginia Department of Conservation and Recreation review). Two
annual fires in a pine-hardwood forest resulted in an
80 percent reduction in Japanese honeysuckle crown
volume and a 35 percent reduction of ground coverage.
While these treatments do not eliminate Japanese
honeysuckle from the site, the authors suggest that
they may reduce the amount of herbicide required
in an integrated management program (Barden and
Matthews 1980).
Fire by itself does not control cogongrass, and in fact
frequent fire promotes cogongrass. Fire can, however,
improve the success of an integrated management
approach using tillage and herbicides. Fire is also important for maintaining native plant diversity in pine
habitat, and restoration of native plant species may be
a critical factor in longer term control of cogongrass
(Howard 2005b).
Controlled burning has not been effective in killing
kudzu, but it can be used to remove vines and leaves
to permit inspection of root crowns for population and
stand monitoring. Fire also promotes seed germination
in kudzu, after which seedlings can be effectively controlled with herbicides. Spring burns are recommended
to reduce soil erosion by winter rainfall (Moorhead
and Johnson 2002, Bugwood Network review). When
removed from a portion of its occupied area, kudzu
can re-invade from water- and bird-disseminated seed
(Brender 1961). Similarly, controlled burning has not
been effective in managing bahia grass because it
sprouts readily after fire (Violi 2000).
Oak-Hickory Woodland Habitat_____
Background
The distribution of oak-hickory woodlands in the
Southeast bioregion has depended on historical fire
management practices. Limited to the most mesic
and protected sites during periods of shortened firereturn intervals, oak-hickory woodland habitats have
increased in area during the fire-exclusion decades of
the early 1900s and continue to occupy many parts of
the Southeast region today (Daubenmire 1978).
Plant communities in this type are dominated by a
variety of oaks and hickories, with a mixture of other
tree species, including maple (Acer spp.), magnolia
(Magnolia spp.), sassafras (Sassafras spp.), and ericaceous shrubs. Several vines commonly occur, including
grape and greenbrier (Smilax spp.). Many pine species
are found in areas with edaphic and/or fire disturbances
(Daubenmire 1978).
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 Fire regimes in oak-hickory habitats are classified
as understory types with return intervals estimated
between 2 and 35 years. Presettlement fire regimes
are poorly understood, although estimates based on
dendrochronology indicate a fire-return interval of 7
to 14 years in the mid-Atlantic region. After European
settlement, fire-return intervals were reduced to 2
to 10 years, with some sites burned annually. At the
present time, the fire regime of oak-hickory forests
is infrequent, low-severity surface fires occurring
principally during spring and fall. They are mainly
human-caused and only burn small areas (Wade and
others 2000).
Oak-hickory woodland habitats in the Southeast are
heavily invaded by aggressive nonnative vines, shrubs,
and trees including kudzu, Japanese honeysuckle,
privet (Ligustrum spp.), bush honeysuckles, and treeof-heaven (Ailanthus altissima). Other invasive species
are found in oak-hickory woodland habitats, but much
less information is available for them. Mimosa (Albizia julibrissin) is a small tree found throughout the
Southeast bioregion in many types of disturbed areas,
including old fields, stream banks, and roadsides (Miller
2003). While it is a common species, little published
information describes its relationship with fire. Giant
reed (Arundo donax) is commonly found in riparian
areas in much of the United States and has been reported as invasive in Georgia, Virginia, and Maryland
(Swearingen 2005). Thorny-olive (Elaeagnus pungens)
is found as an ornamental escape in the Southeast
(Miller 2003), but there are no published reports on the
relationship of this species to fire. Winged euonymus
(Euonymus alatus) is reported in a variety of east coast
habitats, including forests, coastal scrublands and
prairies (USFWS 2004, review), but no information
is available on the relationship of this species to fire.
No information is available concerning a related species, winter creeper (Euonymus fortunei). It is found
in many states throughout the East (Swearingen and
others 2002), but no specific information identifies
habitats where it commonly occurs.
Role of Fire and Fire Exclusion in
Promoting Nonnative Plant Invasions
in Oak-Hickory Woodland
Only limited information is available on the role of
fire and fire exclusion as they affect invasive plant
species in Southeast oak-hickory woodland habitat.
In some cases, fire exclusion seems to promote establishment and spread of nonnatives, while in other
cases the canopy gaps created by fire may increase the
likelihood of establishment and spread of nonnatives.
In more tropical parts of the Southeast, it may be
­assumed that increased light penetration into the plant
105
community will promote establishment and spread of
shade-intolerant nonnative invasive species.
A study of the consequences of hurricane damage in
conservation lands of south Florida demonstrates the
effects of canopy gaps on nonnative plant invasions. Air
potato (Dioscorea bulbifera) and other nonnative vine
species increased after the tree canopy was damaged by
Hurricane Andrew (Maguire 1995). This species may
respond in a similar fashion to canopy gaps created by
fire. Similarly, fire is one of many types of disturbance
that creates canopy gaps that improve the chances for
establishment by tree-of-heaven in old-growth woodland. Additionally, tree-of-heaven seed germination is
delayed and reduced by leaf litter and may therefore
be enhanced by fire when litter is consumed. On the
other hand, fire may produce a flush of herbaceous
growth that could inhibit tree-of-heaven germination
(Howard 2004a, FEIS review). More information is
needed on the effects of fire on seed germination in
this species.
Soil heating may promote kudzu establishment by
scarifying kudzu seedcoats and stimulating germination (Miller 1988). Similarly, mimosa seeds exposed
to fire for 1 to 3 seconds had higher germination rates
than unheated seeds (Gogue and Emino 1979). It has
also been suggested that fire exclusion may promote
Japanese honeysuckle spread (Munger 2002a).
Effects of Plant Invasions on Fuel and Fire
Regime Characteristics in Oak-Hickory
Woodland
The role of invasive plants in altering fire regimes
in the Southeast bioregion is complicated by the existing mosaic of fire exclusion and controlled burning.
Tree-of-heaven, for instance, is found in many types of
woodlands in North America where presettlement fire
regimes have been disrupted in many different ways.
This makes it difficult to make definitive statements
about the potential effect of tree-of-heaven on more
natural fire regimes. The large amount of litter produced by tree-of-heaven from large leaves and broken
branches, and its tendency to form dense thickets, may
contribute to fire spread and crown fires in invaded
areas (Howard 2004a).
A FEIS review speculates that the abundant moist
foliage of kudzu could inhibit fire, effectively lengthening the time between fires in woodland habitats. On
the other hand, the large amount of kudzu biomass
may increase the potential for dormant-season fires
by increasing fuel loads, and its vining nature may
increase the chance of fire crowning. Additionally, increases in standing and surface fuels formed by plants
killed following kudzu invasion may increase both fire
intensity and frequency. The author points out that
106
studies needed to test these hypothetical statements
have not been conducted (Munger 2002b).
Use of Fire to Manage Invasive Plants in
Oak-Hickory Woodland
Fire has not been recommended as a sole management
tool to control tree-of-heaven because of this species’
potential to burn in crown fire, its ability to sprout from
the root crown and/or roots following top-kill from fire,
and the potential for fire to promote seed germination.
Fire has been used to reduce aboveground biomass of
tree-of-heaven (Howard 2004a). A flame-thrower or
weed burning device has been suggested to kill lower
limbs (Hoshovsky 1988, TNC review), but this is not
a population reduction measure.
Fire has been used to reduce cover of Japanese honeysuckle but does not kill plants. Japanese honeysuckle
sprouts from subterranean buds, roots, and stems,
recovering to various levels after fire (Munger 2002a).
Japanese honeysuckle remained a site dominant after
two consecutive annual fires in a pine-hardwood forest in North Carolina (Barden and Matthews 1980).
Experimental plot (abandoned agricultural field) burns
5 years apart near Nacogdoches, Texas, resulted in
Japanese honeysuckle plants with fewer and shorter
prostrate shoots than in unburned plots 1 year after the
last burn, but plants were not killed (Stransky 1984).
Because prostrate shoots are an important part of this
species’ ability to invade native plant communities
(Larson 2000), reduction in numbers of these shoots
could theoretically slow the invasion process.
Seasonality of burns can affect postfire response of
Japanese honeysuckle. Prescribed burns in October
in a Tennessee oak-hickory-pine forest with a maple
and dogwood (Cornus sp.) understory reduced Japanese honeysuckle coverage by 93 percent; burns in
January or March reduced Japanese honeysuckle by
59 percent. Vegetation measurements were taken
at the end of the growing season (September) about
1.5 years after burning (Faulkner and others 1989).
The Nature Conservancy recommends fall, winter, or
early spring prescribed burning to control Japanese
honeysuckle in northern states, when Japanese honeysuckle maintains some leaves and most native plants
are leafless (Nuzzo 1997, TNC review). This improved
ability to target a particular species may have some
applicability in southeastern habitats, but more often
other native species retain leaves through the winter
and may therefore be more subject to damage by fire
at those times.
The Nature Conservancy also suggests that integrating fire and herbicide treatments to control Japanese
honeysuckle may be more effective than either approach
alone, with herbicides applied about a month after
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
sprouting occurs following a late fall or winter burn
(Nuzzo 1997). Application of herbicide about 1 year
after a burn was not effective, possibly because postfire
increases in herbaceous vegetation resulted in less
herbicide contacting Japanese honeysuckle (Faulkner
and others 1989). Fire is also helpful in controlling
fire-intolerant Japanese honeysuckle seedlings and
young plants. Efforts should be made to avoid soil
disturbance as much as possible to reduce subsequent
germination of Japanese honeysuckle seeds in the seed
bank (Nuzzo 1997).
Prescribed fires have been suggested for controlling
bush honeysuckles (Tatarian honeysuckle, Morrow’s
honeysuckle, Bell’s honeysuckle (Lonicera X bella),
and Amur honeysuckle) in fire-adapted communities
(Nyboer 1990, Illinois Nature Preserves Commission
review). Spring burns kill bush honeysuckle seedlings
and top-kill mature plants; however, plants sprout
readily after fire. Effective control may come from
annual or biennial fires conducted for 5 years or more
(Nyboer 1990).
It has been suggested that Chinese privet is intolerant of fire (Matlack 2002) and can be managed
successfully by repeated fire, especially on sites with
low stem density and high fine fuel loads (Batcher
2000a, TNC review). A single fire does not result in
sufficient kill of mature plants (Faulkner and others
1989) but instead promotes sprouts from root crowns
and/or roots (Munger 2003c, FEIS review). Chinese
privet burns poorly without additional fuel. However,
if sufficient low-moisture fuels are available (Batcher
2000a), annual fires may substantially reduce or kill
aboveground portions of Chinese privet, although
they will not eliminate it from a site. Three annual
prescribed burns did not eradicate Chinese privet from
areas where fire had been excluded for more than 45
years (Munger 2003c). Platt and Stanton (2003) suggest
that dominance by Chinese privet cannot be reversed,
but increases in population size can be prevented with
short return interval, lightning-season fires. Japanese
privet (Ligustrum japonicum) and European privet
(L. vulgare) also occur in this vegetation type, but no
specific information on management and fire for these
species in this vegetation type is available.
Prescribed burns have been suggested as part of a
strategy to manage kudzu. Information on the limitations of prescribed fire and effects of kudzu removal
on native vegetation (presented in the “Pine and Pine
Savanna Habitat” section page 100) is relevant to this
habitat as well.
Air potato (Dioscorea bulbifera) invades woodland
habitats throughout Florida (Schmitz and others 1997).
While only a limited amount of research has been conducted, prescribed fire may be useful in killing stem
growth (Morisawa 1999b, TNC review) and bulbils
(Schultz 1993, TNC review) of air potato in woodlands.
A related species, Chinese yam (Dioscorea oppositifolia), has been reviewed by The Natural Conservancy
(Tu 2002b). Chinese yam is found in mesic bottomland
forests, along streambanks and drainageways in many
states of the Southeast. Only very limited information
is available about the use of fire to manage this species. It was noted that reduced amounts of Chinese
yam were present the year following a fall wildfire in
Great Smoky Mountains National Park (Tu 2002b),
but the specific habitat information is not available.
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 107
Tropical Hardwood
Forest Habitat_ __________________
Background
Scattered throughout the grassland and savanna
plant communities of south Florida are “islands” of
tropical hardwood species, often called hammocks,
and typically found on somewhat drier sites. Common
species include gumbo limbo (Bursera simaruba), black
ironwood (Krugiodendron ferreum), inkwood (Exothea
paniculata), lancewood (Ocotea coriacea), marlberry
(Ardisia escallonoides), pigeon plum (Coccoloba
diversifolia), satinleaf (Chrysophyllum oliviforme),
poisonwood (Metopium toxiferum), and white stopper
(Eugenia axillaris). While limited in extent compared
to the other habitats discussed in this chapter, they are
important because of the plant diversity they provide
within the grassland landscape.
Fire regime in this habitat varies from low-intensity
surface fires to crown fires, with an estimated presettlement return interval of 35 to over 200 years. Fire
has not been the dominant force in shaping hardwood
hammock plant communities because they are usually
difficult to burn. Although many of the hardwood species sprout following top-kill from fire, the stands can
be destroyed by fire during periods of drought if the
organic soil is consumed (Myers 2000).
Many hardwood hammocks are being aggressively
invaded by air potato, melaleuca (fig. 6-10), and Old
World climbing fern. Water yam (Dioscorea alata) has
been reported in coastal hammocks in Florida (FLEPPC
1996). This species is related to air potato and may act
in a similar manner to air potato in relation to fire,
although no specific information is available for either
species.
Role of Fire and Fire Exclusion in
Promoting Nonnative Plant Invasions in
Tropical Hardwood Forest
Limited information is available regarding the role
of fire in promoting nonnative plant invasions in
tropical hardwood hammocks. Melaleuca is capable
Figure 6-10—Melaleuca saplings (green trees) ringing a Cypress clump at the
­Loxahatchee National Wildlife Refuge around 1994. The melaleuca eventually overtook the cypress (Ferriter, personal communication 2007). (Photo by Amy Ferriter,
South Florida Water Management District, Bugwood.org.)
of spreading quickly into this habitat following fires
that remove most of the vegetation and expose bare
mineral soil (Bodle and Van 1999).
Old World climbing fern is rapidly invading hardwood
hammocks, but it is not known to what extent invasion
is dependent on or retarded by fire. Anecdotal evidence
suggests that the fern appears to grow especially well
in hammocks where native vegetation was either damaged or killed by fire during a drought. Researchers
and natural area managers in southern Florida suspect
that Old World climbing fern is particularly robust in
areas where native tree island vegetation was damaged
or killed by a fire during the drought of 1989–1990. It
has also been suggested, although not demonstrated,
that convection currents generated by burning fern
growth that has formed a trellis up into tall trees could
increase the dispersal of spores (Ferriter 2001). Old
World climbing fern plants sprout and recover rapidly
after low-severity fire (Stocker and others 1997).
The influence of melaleuca and Old-World climbing
fern on fuels and fire regimes has been discussed in
previous parts of this chapter, as has the use of fire for
managing these species. Fire by itself is not an effective
method to manage melaleuca or Old World climbing
108
fern, although suggestions have been made for ways in
which fire could be incorporated into an integrated management approach with other control techniques. Since
native species in this habitat probably did not evolve in
a regime of frequent fire, increasing fire frequency with
prescribed burning might have unintended effects on
native plant species (Ferriter 2001).
Cypress Swamp Habitat___________
Very limited information is available about the
relationship between fire and invasive species in
cypress swamp habitat. Depressional wetlands in
central and south Florida dominated by bald cypress
(Küchler’s (1964) Southern Floodplain Forest) had a
presettlement stand-replacement fire-return interval
estimated at 100 to 200 years or greater (Wade and
others 2000). This fire regime has been altered by
invasions of melaleuca and Old World climbing fern.
Both species increase the probability of more frequent
stand-replacement fires because of the ease with which
they burn and because Old World climbing fern can
form a fuel bridge between adjacent, more frequently
burned habitats and the much wetter cypress swamp
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
(Langeland 2006). Information on the role of these
species in relation to fire is also presented in the “Wet
Grassland Habitat,” “Pine and Pine Savanna Habitat”
and “Oak-Hickory Woodland Habitat” sections.
Catclaw mimosa (Mimosa pigra) occurs on about
1,000 acres (400 ha) in Florida, including cypress
swamp habitat. While fire has been used experimentally in Australia to clear mature plants, enhance seed
germination for subsequent herbicide application, and
kill some seeds (Lonsdale and Miller 1993), no similar
studies have been conducted in the United States.
Conclusions and Summary_ _______
The importance of fire and fire management in
influencing species composition and dynamics of
plant communities in the Southeast bioregion is often
stated, if not completely understood. Incomplete or
contradictory information, for instance, describes the
plant communities prior to human influence (Stanturf
and others 2002). With its high number of lightning
strikes, its many fire-adapted communities, and its
historical human dependence on fire-maintained
habitats, fire has probably been a more important
factor in the Southeast than in any other broadly defined region of the country. Plant communities such
as longleaf pine once covered millions of acres when
human populations supplemented naturally occurring
fires with intentional blazes (Chapman 1932). Even
wet grasslands in the Southeast tend to be flammable
and are adapted to frequent fires (Leenhouts 1982;
Schmalzer and others 1991; Wade 1988; Wade and
others 1980).
Human influence on fire regime was accompanied
by a large number of intentional introductions of
plants for agricultural, horticultural, medicinal, and
religious purposes, as well as many accidental imports.
The large number of intentional introductions and
the escape of these introduced plants led to very high
levels of invasive plants in the Southeast bioregion,
especially Florida. More than 25,000 species and cultivars have been introduced to Florida (D. Hall, personal
communication, cited by Gordon 1998), a state with
around 2,523 native species (Ward 1990). While many
of these introductions have served their intended purposes, a small proportion (about 10 percent (Gordon
1998)) has caused unintended damage to forest and
conservation lands. It is not clear whether the large
number of nonnative plant invasions in this bioregion
is due to the large number of introductions (in other
words, propagule pressure), or whether habitats in the
Southeast are more susceptible to invasion. The high
number of plant species introductions in the Southeast
may be related to the diversity of cultural origins of
the human populations and a range of climates from
tropical to temperate. This latter factor could also
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 influence the susceptibility of the region by providing
a wider range of potentially suitable conditions for
establishment and spread.
The causes and effects of invasive plant establishment
and spread are related to fire and fire management to
varying degrees. The role of fire and fire exclusion in
promoting nonnative plant invasions is very different
among the five habitats discussed in this chapter. Fire
exclusion policy, of course, does not mean the absence
of fire. Historical and current efforts to exclude fire
have often led to less frequent but more severe fires,
which then can lead to substantial changes in native
and invasive species populations (Wade and others
2000).
Role of Fire and Fire Exclusion in
Promoting Nonnative Plant Invasions
The complex and dramatic relationship between
invasive species and both fire and fire exclusion in
the Southeast bioregion may be best exemplified by
melaleuca invasion in pine, wet grassland, and tropical hardwood habitats in south Florida. Melaleuca has
spread into thousands of acres apparently without need
of fire, but has spread most dramatically following
natural fire, controlled burns, and uncontrolled fire
following decades of fire exclusion efforts.
Wet grasslands may be the habitat type most affected
by fire and fire exclusion. Expansion of melaleuca following fire has been more thoroughly reported for wet
grassland habitats (Ferriter 1999). Fire in wet grasslands is also possibly responsible for increases in Old
World climbing fern (Langeland 2006) and chinaberry
(Menvielle and Scopel 1999; Tourn and others 1999),
although the evidence is principally anecdotal. Successful exclusion of fire has been blamed for Brazilian
pepper invasions of salt marshes (Schmalzer 1995)
and for Chinese tallow (Bruce and others 1995) and
tall fescue invasions (Eidson 1997) in wet grassland
habitat.
Fire exclusion has clear ecological impacts in pine
habitats, which frequently accumulate additional
woody shrub species in the absence of fire and ultimately become hardwood dominated habitats with
a minor pine component (DeCoster and others 1999;
Heyward 1939; Platt 1999; Slocum and others 2003;
Streng and others 1993; Walker and Peet 1983). Less
has been studied, however, about these longer-term
changes and the interplay of fire and invasives. Brazilian pepper (Loope and Dunevitz 1981) and Japanese
honeysuckle (Munger 2002a) are among the woody
shrubs that are promoted in pine habitats when fire
is excluded. Conversely, cogongrass in pine habitats is
promoted by frequent surface fires and, in fact, requires
some regular disturbance to maintain its dominance
(Howard 2005b).
109
Effects of Plant Invasions on Fuel and Fire
Regime Characteristics
Some of the most obvious visual changes in fire occur when melaleuca invasions fuel crown fires in wet
grassland communities. But while these conflagrations
are dramatic, the removal of aboveground plant material is not a substantial change to the wet grassland
fire regime. It is the secondary effects of these fires,
such as consumption of the surface of organic soils
followed by flooding, that can lead to major changes
in plant communities (Wade and others 2000).
Some evidence suggests that when species such as
Chinese privet, Chinese tallow, and Brazilian pepper
replace native plant species in wet grasslands, fine
fuel loads and horizontal continuity are reduced (for
example, Doren and Whiteaker 1990; Doren and others
1991; Grace 1999; Platt and Stanton 2003). Change
in fine fuels may lead to reduced fire frequency and
intensity, and increased fire patchiness. There is,
however, little experimental evidence to support these
suggestions.
The most substantial changes in fire regime caused by
nonnative plant invasion are probably in pine habitats.
Invasions of Old World climbing fern, and possibly its
congener Japanese climbing fern, increase incidence
of crown fires, carry fire across wetland barriers that
would have stopped the fire if they had not contained
Old World climbing fern, and possibly “kite” fire to new
locations (Langeland 2006). Cogongrass changes fire
behavior and effects in pine habitats. Cogongrass invasions lead to increased biomass, horizontal continuity,
and vertical distribution of fine fuels when compared
with uninvaded pine savanna, and higher maximum
fire temperatures have been reported (Lippincott
2000). In this particular example, the detailed studies
necessary to show actual replacement of native species have not been conducted. Melaleuca changes the
fire regime in pine stands from frequent, low-severity
surface fires to a mixed fire regime with less frequent
fires and greater incidence of crown fires. These crown
fires are often lethal to pines but not to melaleuca
(Myers 2000). It is possible that Brazilian pepper invasions have opposite effects and reduce fire intensity
and fire spread where it is densely distributed (Doren
and Whiteaker 1990; Doren and others 1991).
Use of Fire to Manage Invasive Plants
Among the habitats covered in this section, pine and
pine savanna habitats are probably the most conducive to use of fire to manage invasive plants, in large
part because of the role of fire in maintaining these
pyric communities. Controlled burns in oak-hickory
woodland have resulted in reductions of Japanese
honeysuckle (Barden and Matthews 1980; Williams
1994) and reduced spread of Chinese privet (Batcher
110
2000a). Since frequent prescribed fire in oak-hickory
woodland will favor pine species at the expense of
young hardwoods, it will probably be more difficult
to maintain as aggressive a burning practice in oakhickory woodland than in pine and pine savanna.
In the presence of propagule sources from invasive
species such as melaleuca and Old World climbing fern,
prescribed fire in wet grasslands must be conducted
with extreme care and may result in expansion of these
species. The margin between use of fire for successful
reduction of melaleuca seedlings following a seedrelease event and accidental expansion of melaleuca
into fire-cleared seed beds is very small and is affected
by weather and water management patterns out of the
manager’s control (Belles and others 1999).
Additional Research Needs
Among the general information needs related to fire
and invasive plants, several specific needs stand out.
While we have case studies on short-term effects of
various fire related practices for individual species, we
don’t know which fire management practices in which
habitats will provide the most effective means of reducing
existing invasive plant populations or preventing future
invasions. This is not an easy area of research, in part
because we do not know which of the tens of thousands
of novel species that could be introduced to southeastern
habitats will become management problems, making it
nearly impossible to know what practices will provide the
most future benefit. There are many nonnative species
already present in Southeast bioregion habitats for which
no ­information is available. Included in this category
are giant reed, field bindweed (Convolvulus arvensis),
Chinese silvergrass (Miscanthus sinensis), princesstree
(Paulownia tomentosa), golden bamboo (Phyllostachys
aurea), multiflora rose (Rosa multiflora), sericea lespedeza, five-stamen tamarisk (Tamarix chinensis), French
tamarisk (T. gallica), smallflower tamarisk (T. parviflora), saltceder (T. ramosissima), bigleaf periwinkle
(Vinca major), common periwinkle (V. minor), Japanese
wisteria (Wisteria floribunda), and Chinese wisteria
(Wisteria sinensis).
At a minimum we need longer-term studies that
document broad species changes in population size
and distribution. For instance, it has been suggested
that replacement of wet grassland species by invasive
hardwood shrubs (Brazilian pepper) and trees (Chinese
tallow) results in reduced fine fuel load and horizontal
continuity (Doren and Whiteaker 1990; Doren and
others 1991; Grace 1999). These fuel changes could
logically lead to changes in fire frequency, severity,
and patchiness, but that has yet to be documented.
There is also potential for subsequent changes in native plant species coverage and/or diversity—a topic
that deserves further study.
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008
Because so much of the Southeast is relatively lowlying, the effect of natural and artificially-manipulated
hydrology on the relationship between fire and invasive
plants needs to be examined. For instance, we know
that artificially lowered surface-water elevations in
south Florida wet grasslands lead to increased fires
and exposure of mineral soil that facilitate melaleuca
invasions, but management practices that could prevent such invasions have not been determined.
Emerging Issues
The Southeast has more fire and more invasive
plants than most other parts of the country. To further
complicate the situation, the Southeast is also rapidly
increasing its human population. The “sunbelt” is the
fastest growing part of the United States, with a 21
percent increase in population between 1970 and 1980,
and an 18 percent increase between 1980 and 1997
(NPA 1999). It remains to be seen whether prescribed
fire practices can be implemented and maintained
with more and more urban incursions into forest and
natural areas.
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008 Global warming and related increases in carbon dioxide may well affect the relationships among invasive
plants, native plant communities, and fire. Lightning
frequency is expected to increase with global warming.
The U.S. Global Change Research Program’s National
Assessment Synthesis Team has predicted that the
“seasonal severity of fire hazard” will increase about
10 percent for much of the United States, but a 30
percent increase in fire hazard is predicted for the
Southeast (NAST 2000).
Major changes in plant communities have occurred
in the Southeast because of the interaction of invasive
nonnative plants and fire management policy and
practice. However, it is not a given that the Southeast
would have avoided its serious invasive plant problems
if presettlement fire regimes had been maintained,
nor that reinstating presettlement fire regimes in the
1800s would have prevented problems during the next
two centuries.
Better understanding of the relationship between
invasive plants, fire, and fire management is certainly
needed if resource managers are to maximize their
ability to prevent further nonnative plant invasions
and adjust fire policy to both reduce the existing invasive plant populations and achieve other management
objectives.
111
Notes
112
USDA Forest Service Gen. Tech. Rep. RMRS-GTR-42-vol. 6. 2008